1 2011-08-19 00:00:11 <CIA-101> bitcoin: Con Kolivas * raaa2e19e0e08 cgminer/phatk110816.cl:
  2 2011-08-19 00:09:10 <Diablo-D3> thaaaat cant be a good idea.
  3 2011-08-19 00:09:24 <Diablo-D3> seeing as branches are more expensive than jumps
  4 2011-08-19 00:55:01 <zeropointo> anyone here ever try boost asio for cross platform development?
  5 2011-08-19 01:43:29 <diki> So, i want to ask. I currently have the idea of making an exchange like doublec's i.e a nmc<->bitcoin,i0coin<->bitcoin and so forth exchange. the problem is i have no idea about the algorithms. I.e how do you calculate the price etc
  6 2011-08-19 01:43:56 <diki> i know that the buy price must always be higher than the sell price even if it's just by one satoshi
  7 2011-08-19 01:44:02 <diki> but what is the algorithm?
  8 2011-08-19 01:49:53 <doublec> there is no price algorithm
  9 2011-08-19 01:50:07 <doublec> people list what they're buying at, others list what they're selling at
 10 2011-08-19 01:50:09 <diki> ofc there is
 11 2011-08-19 01:50:34 <diki> and the algorithm is what makes the price
 12 2011-08-19 01:50:36 <gjs278> there is no price algorithm
 13 2011-08-19 01:50:37 <gjs278> no
 14 2011-08-19 01:50:41 <gjs278> the people are what makes the price
 15 2011-08-19 01:50:45 <gjs278> you set your own price on mtgox
 16 2011-08-19 01:50:49 <diki> and how do you calculate it based on that?
 17 2011-08-19 01:50:54 <gjs278> my head
 18 2011-08-19 01:51:01 <gjs278> I figure that coins might be worth $9
 19 2011-08-19 01:51:04 <gjs278> so I put them up for $9
 20 2011-08-19 01:51:17 <gjs278> if people enter buy orders for $9 then they buy my coins
 21 2011-08-19 01:51:24 <diki> erm, ok, i know the user can input any price
 22 2011-08-19 01:51:28 <gjs278> yes
 23 2011-08-19 01:51:30 <doublec> the 'price' is just the price that the last buy/sell occured at
 24 2011-08-19 01:51:33 <gjs278> so what does the system have to do
 25 2011-08-19 01:51:36 <gjs278> besides record that
 26 2011-08-19 01:51:38 <gjs278> that's it
 27 2011-08-19 01:51:39 <coderrr> diki, please dont make an exchange
 28 2011-08-19 01:51:56 <diki> coderrr:thats like asking mtgox to stop trading
 29 2011-08-19 01:52:12 <coderrr> no its liek asking u to not make an exchange
 30 2011-08-19 01:52:12 <gjs278> lol
 31 2011-08-19 01:52:30 <doublec> no, it's like asking mtgox not to make nuclear power station software
 32 2011-08-19 01:53:17 <doublec> diki: work out some orders, buys and sells on paper and you'll see how it works
 33 2011-08-19 01:54:49 <diki> doublec what is your fee?
 34 2011-08-19 01:54:58 <doublec> diki: fee for what?
 35 2011-08-19 01:54:59 <diki> i am not asking so i can copy it
 36 2011-08-19 01:55:08 <diki> fee per trade
 37 2011-08-19 01:55:16 <doublec> diki: it's on the front page of the exchange
 38 2011-08-19 01:55:35 <diki> it says 0.2%
 39 2011-08-19 01:55:43 <diki> i honestly dont know what that is in satoshis
 40 2011-08-19 01:55:50 <doublec> it's a percentage
 41 2011-08-19 01:55:59 <doublec> therefore it's not measured in satoshis
 42 2011-08-19 01:56:12 <diki> convert it to bitcoins then
 43 2011-08-19 01:56:19 <doublec> are you trolling?
 44 2011-08-19 01:56:23 <diki> i am not
 45 2011-08-19 01:56:29 <doublec> I can't conceive of why that question would be asked
 46 2011-08-19 01:56:36 <diki> i have no idea what 0.2% of XX price is
 47 2011-08-19 01:56:50 <gjs278> diki you noob
 48 2011-08-19 01:56:55 <gjs278> I will fire a rocketlauncher at you
 49 2011-08-19 01:57:42 <diki> is 0.2% 0.002 per ?coin trade?
 50 2011-08-19 01:58:58 <diki> or is it 0.02?
 51 2011-08-19 01:59:27 <doublec> it is 0.2%
 52 2011-08-19 01:59:34 <diki> that tells me nothing really
 53 2011-08-19 01:59:44 <diki> how much ?coins are taken from the balance per trade?
 54 2011-08-19 01:59:49 <diki> that is what i am asking
 55 2011-08-19 01:59:52 <doublec> 0.2%
 56 2011-08-19 02:01:02 <diki> so basically if 200 ?coins are traded, then 0.2% fee of them coins will be taken from the balance?
 57 2011-08-19 02:01:15 <diki> which means if it was 1 coin then 0.2% of that single coin>
 58 2011-08-19 02:01:36 <doublec> yes
 59 2011-08-19 02:01:59 <diki> so the higher coins per trade, the more you take, tho still 0.2%
 60 2011-08-19 02:02:38 <diki> but even if you traded in small amounts, in the long run you still take the same amount as you would with more coins per trade
 61 2011-08-19 02:03:49 <doublec> right
 62 2011-08-19 02:04:20 <diki> Ok, what would happen to an exchange if there were no fees at all. I mean apart from the exchange not earning at all, would it affect the trading in some way?
 63 2011-08-19 02:04:35 <diki> would it be abused somehow?
 64 2011-08-19 02:04:53 <doublec> there are exchanges out there with no fees
 65 2011-08-19 02:04:57 <doublec> eg. intersango.us
 66 2011-08-19 02:05:16 <diki> this intersango.us is just like britcoin
 67 2011-08-19 02:05:57 <doublec> it uses the same software
 68 2011-08-19 02:06:05 <doublec> it's open source
 69 2011-08-19 02:06:34 <diki> Ok, so how do you keep the buy price higher than the sell price?
 70 2011-08-19 02:06:45 <diki> even if it's by 0.0000001 higher than sell
 71 2011-08-19 02:07:15 <doublec> the buy won't be higher than the sell, because if it was, the sell would be sold
 72 2011-08-19 02:07:45 <diki> hmm? but on any exchange the sell price is always lower than the buy price
 73 2011-08-19 02:07:58 <diki> i.e buying a bitcoin is 10.96 but selling a bitcoin is 10.90 per coin
 74 2011-08-19 02:08:42 <doublec> perhaps our terminology is different
 75 2011-08-19 02:09:04 <noagendamarket> if theres no fee youd get bots trading
 76 2011-08-19 02:09:33 <diki> people like no fee stuff
 77 2011-08-19 02:09:55 <doublec> by "buy" i refer to 'buy orders' from exchange users
 78 2011-08-19 02:10:06 <doublec> and "sell" i refer to 'sell orders' from exchange users
 79 2011-08-19 02:10:07 <diki> oh, so your buy is actually sell
 80 2011-08-19 02:10:48 <diki> yes, this terminology  is confusing
 81 2011-08-19 02:10:58 <diki> in fact, the mtgox ticker is the same way
 82 2011-08-19 02:11:49 <doublec> your thinking of a "money exchange" i think that doesn't match orders from users
 83 2011-08-19 02:11:54 <doublec> but rather buys/sells itself
 84 2011-08-19 02:12:36 <diki> no i am not thinking of a money exchange
 85 2011-08-19 02:12:43 <diki> simply ?coin trading
 86 2011-08-19 02:12:48 <doublec> in that case the price the user buy scurrency at will be higher than that which they can sell currency at
 87 2011-08-19 02:12:51 <diki> no ?coin for <fiat> currency
 88 2011-08-19 02:13:50 <diki> so if this intersango is open source, i can use it's "engine" right?
 89 2011-08-19 02:14:16 <doublec> depending on the license. I don't know what it is.
 90 2011-08-19 02:15:22 <diki> so, should i assume you've made like >100btc per 100k trades on any of the exchanges?
 91 2011-08-19 02:19:32 <doublec> nope
 92 2011-08-19 02:19:39 <doublec> the i0 exchange made 3btc
 93 2011-08-19 02:20:10 <diki> something doesnt sound right
 94 2011-08-19 02:20:18 <doublec> that would be your maths
 95 2011-08-19 02:20:48 <diki> so, if there were 100k trades, each with 10k coins, only then would you have made >3btc?
 96 2011-08-19 02:21:43 <doublec> let's say 100,000 coins are traded. At 0.003 btc per coin. That's 300 btc.
 97 2011-08-19 02:21:56 <doublec> 0.2% of that is 0.6 btc
 98 2011-08-19 02:22:39 <diki> please do share how you calculate that with the percentage
 99 2011-08-19 02:22:44 <diki> its always been a mystery to me
100 2011-08-19 02:22:45 <doublec> I just did
101 2011-08-19 02:23:08 <diki> i know that 50% of 100k is 50k
102 2011-08-19 02:23:14 <diki> as for 0.2% i have no clue
103 2011-08-19 02:23:35 <doublec> divide the percentage by 100
104 2011-08-19 02:23:44 <doublec> so 100,000 * (50 / 100) = 50,000
105 2011-08-19 02:23:59 <doublec> so 300 * (0.2 / 100) = 0.6
106 2011-08-19 02:26:43 <doublec> diki: makes sense now?
107 2011-08-19 02:26:51 <diki> well, this is what happens when you have ADD and dont pay attention in math class
108 2011-08-19 02:27:02 <diki> add=attention deficit disorder
109 2011-08-19 02:30:59 <noagendamarket> please dont start an exchange if you dont know how
110 2011-08-19 02:31:16 <diki> this is why i am learning
111 2011-08-19 02:31:22 <diki> i believe there is a saying
112 2011-08-19 02:31:23 <noagendamarket> weve had enough "weekend projects"
113 2011-08-19 02:31:25 <diki> no one is born taught
114 2011-08-19 02:31:39 <noagendamarket> mybitcoin
115 2011-08-19 02:31:58 <diki> myb0tfailcoin is not of my concern
116 2011-08-19 02:32:08 <diki> that guy waited for 1 confirm
117 2011-08-19 02:32:28 <diki> in fact, he prolly didnt even wait for a confirm(who knows)
118 2011-08-19 02:37:55 <CIA-101> bitcoinjs/bitcoinjs-gui: Stefan Thomas master * rc6040d5 / (scripts/phone/index.js stylesheets/phone/style.css): Added welcome screen. - http://bit.ly/ntkknu
119 2011-08-19 02:40:11 <CIA-101> libbitcoin: genjix * r9f2c3f944b09 /doc/reorganize.py: Added Python reference implementation of block chain reorganizer for postgresql_storage module.
120 2011-08-19 02:40:13 <CIA-101> libbitcoin: genjix * rdfea87570170 / (examples/poller.cpp src/storage/postgresql_storage.cpp): BUGFIX: LOL service was calling itself recursively forever XD
121 2011-08-19 03:11:26 <vsrinivas> when does a client rebroadcast a transaction?
122 2011-08-19 03:18:43 <sgornick> vsrinivas: it is a random interval ... can be a half hour, sometimes more even.
123 2011-08-19 03:47:27 <shadders> pushpool rewrites difficulty to :7fffff0000000000000000000000000000000000000000000000000000000000
124 2011-08-19 03:48:16 <shadders> slush pool rewrites to : ffffffffffffffffffffffffffffffffffffffffffffffffffffffff00000000
125 2011-08-19 03:48:49 <neofutur> which means one of them is easyer ?
126 2011-08-19 03:48:54 <shadders> anyone know why slush uses a lower difficulty?
127 2011-08-19 03:49:28 <shadders> I put slush difficulty into poolserverj because I incorrectly asssumed it was difficulty 1
128 2011-08-19 03:50:29 <shadders> neofutur: slush's should be easier... it's a higher target, easir to get a hash below it
129 2011-08-19 03:54:11 <neofutur> thanks for the news, interesting
130 2011-08-19 03:54:47 <neofutur> a good column to add on https://en.bitcoin.it/wiki/Comparison_of_mining_pools
131 2011-08-19 03:55:29 <neofutur> how do you find this info for other pools ?
132 2011-08-19 03:58:51 <shadders> send a getwork request and look at the response "target" field
133 2011-08-19 04:00:32 <imsaguy> shadders, the lower difficulty means slower clients can still do 'shares'
134 2011-08-19 04:01:01 <imsaguy> eventually one of the easier difficulty hashes will be enough to qualify as a current difficulty hash
135 2011-08-19 04:02:02 <shadders> it would be increasing his server load massively
136 2011-08-19 04:02:17 <luke-jr> shadders: gosh, should you be writing a pool server when you don't know these basics? XD
137 2011-08-19 04:02:28 <luke-jr> but pushpool uses ffffffffffffffffffffffffffffffffffffffffffffffffffffffff00000000 too fwiw
138 2011-08-19 04:02:48 <imsaguy> lol
139 2011-08-19 04:02:56 <imsaguy> yeah, he'll be dealing with a lot of hashes
140 2011-08-19 04:03:03 <imsaguy> but thats what allows slower machines to participate
141 2011-08-19 04:03:18 <imsaguy> you're missing the very premise of it.
142 2011-08-19 04:03:33 <shadders> sorry, will go study... maybe while I'm away you can make pushpool a bit faster :p
143 2011-08-19 04:05:13 <shadders> yes I get that...  just wondering why they are diffeent...
144 2011-08-19 04:05:29 <shadders> luke-jr: this is from pushpool: {"id":1,"error":null,"result":{"midstate":"126a76e126f390b90611962782af8a342c5796a881b23b20b59ae7f0f857c533","target":"7fffff0000000000000000000000000000000000000000000000000000000000","data":"0000000113389092f6e837be3a6dc96c0cc78abffe11f9c9ff4f57a6000007010000000060ceb36a8b15b2f428d49264b6f0f8ab3739a2559044248f46579be3e3ad11404e4df8601a094a86000000000000008000000000000000000000000000000000000000000000000
145 2011-08-19 04:05:30 <shadders> 00000000000000000000000000000000080020000","hash1":"00000000000000000000000000000000000000000000000000000000000000000000008000000000000000000000000000000000000000000000000000010000"}}
146 2011-08-19 04:05:57 <luke-jr> shadders: you misconfigured pushpool I bet then
147 2011-08-19 04:06:49 <shadders> # rewrite returned 'target' to difficulty-1?
148 2011-08-19 04:06:50 <shadders> 
149 2011-08-19 04:07:27 <shadders> other than that it's stock 0.5.1 with JK's 0.99patch
150 2011-08-19 04:09:01 <shadders> luke-jr: what have I done wrong?
151 2011-08-19 04:09:40 <luke-jr> JoelKatz seems annoying; why can't he make proper branches and send pull reqs
152 2011-08-19 04:10:06 <luke-jr> probably something in his patch
153 2011-08-19 04:10:07 <shadders> sorry not with JK's patch... that was for bitcoind..
154 2011-08-19 04:10:12 <luke-jr> o
155 2011-08-19 04:10:22 <luke-jr> no idea what you're doing wrong then
156 2011-08-19 04:10:31 <luke-jr> that target you're giving is basically impossibl
157 2011-08-19 04:11:04 <shadders> that's difficulty 1
158 2011-08-19 04:11:12 <luke-jr> nope
159 2011-08-19 04:11:15 <shadders> according to https://en.bitcoin.it/wiki/Target
160 2011-08-19 04:12:06 <luke-jr> diff1 = 000000000000000000000000000000000000000000000000ffff000000000000
161 2011-08-19 04:13:37 <shadders> sorry yr gight it was a messed up version of pp... too many versions installed, lose track of which one I'm using
162 2011-08-19 04:16:00 <shadders> no wonder I buggered up the basics when I'm learning of a wiki that's wrong!
163 2011-08-19 04:16:25 <luke-jr> wiki isn't wrong, just big endian
164 2011-08-19 04:16:40 <shadders> oh ffs
165 2011-08-19 04:23:50 <shadders> so is ffffffffffffffffffffffffffffffffffffffffffffffffffffffff00000000
166 2011-08-19 04:24:01 <shadders> little endian then?
167 2011-08-19 04:24:08 <luke-jr> no, it's big endian
168 2011-08-19 04:24:12 <luke-jr> wait
169 2011-08-19 04:24:23 <luke-jr> no
170 2011-08-19 04:24:27 <luke-jr> that's little endian, right
171 2011-08-19 04:24:52 <luke-jr> it's also the mixed-endian used in getwork requests
172 2011-08-19 04:25:10 <shadders> you're messing with my head now... I'm gonna ask satoshi
173 2011-08-19 04:25:18 <luke-jr> diff1 LE = 0000000000000000000000000000000000000000000000000000ffff00000000
174 2011-08-19 04:25:28 <luke-jr> diff1 xE = 000000000000000000000000000000000000000000000000ffff000000000000\n1576529
175 2011-08-19 04:25:54 <luke-jr> shadders: lol
176 2011-08-19 04:26:23 <shadders> do I even wanna know.. what? x endian?
177 2011-08-19 04:26:46 <luke-jr> miXed endian
178 2011-08-19 04:26:48 <luke-jr> used by getworks
179 2011-08-19 04:27:29 <luke-jr> mixed endian = &89abcdef01234567
180 2011-08-19 04:28:16 <shadders> wtf???
181 2011-08-19 04:28:44 <shadders> I seriously nearly gave up the psj project when I cam across all this endian shite...
182 2011-08-19 04:29:02 <luke-jr> :D
183 2011-08-19 04:29:23 <shadders> I still don't know what benefit there is to it...
184 2011-08-19 04:29:42 <luke-jr> none whatsoever
185 2011-08-19 04:29:58 <shadders> so why not pick one and stick to it?
186 2011-08-19 04:31:25 <luke-jr> because Satoshi likes to piss people off with random crap
187 2011-08-19 04:31:27 <erska> to make a compromise between the two, so that the other endianness does not feel rejected and lonely ;)
188 2011-08-19 04:34:29 <shadders> probably why he pissed off... was afraid some programmer was going to lose the plot trying to deal with endians and try to kill him.
189 2011-08-19 04:35:00 <luke-jr> lol
190 2011-08-19 04:35:27 <luke-jr> he probably had a nice long todo list of bugs he found and was gonna fix before doing the real thing
191 2011-08-19 04:35:32 <luke-jr> and realized it was too late
192 2011-08-19 04:36:14 <shadders> "Numerous other orderings, generically called middle-endian or mixed-endian, are possible. On the PDP-11 (16-bit little-endian) for example, the compiler stored 32-bit values with the 16-bit halves swapped from the expected little-endian order. This ordering is known as PDP-endian."
193 2011-08-19 04:37:21 <shadders> words fail me... ok so what's the definition of the mixed endian bitcoin uses?  or, let me guess... there's a few of them
194 2011-08-19 04:38:53 <luke-jr> shadders: overall little-endian split into 32-bit big-endian chunks
195 2011-08-19 04:39:12 <shadders> there should be victim support group for this
196 2011-08-19 04:39:20 <shadders> yr it luke-jr
197 2011-08-19 04:39:35 <luke-jr> I torture my victims.
198 2011-08-19 04:40:29 <cjdelisle> so 8 little endian integers which themselves are arranged in big endian order?
199 2011-08-19 04:43:36 <shadders> so... FFAA0000 LE should be 0000FFAA or 0000AAFF in xE?
200 2011-08-19 04:45:10 <luke-jr> 0000AAFF I think
201 2011-08-19 04:45:22 <luke-jr> cjdelisle: you have the endians backward there
202 2011-08-19 04:45:36 <luke-jr> cjdelisle: the 32-bit blocks are big-endian; they are arranged in little-endian
203 2011-08-19 04:45:46 <cjdelisle> ahh ok
204 2011-08-19 04:46:02 <cjdelisle> must have been because ntohl() was easy to use? o_O
205 2011-08-19 04:46:23 <luke-jr> cjdelisle: NFC why
206 2011-08-19 04:46:29 <luke-jr> someone must have been drunk or something
207 2011-08-19 04:46:33 <cjdelisle> heh
208 2011-08-19 04:49:28 <shadders> well that must be why my share validation is broken.. parsing ffffffffffffffffffffffffffffffffffffffffffffffffffffffff00000000 as big-endian
209 2011-08-19 05:30:09 <CIA-101> bitcoin: Con Kolivas * rbd79a61c439c cgminer/ (findnonce.h poclbm110816.cl): Move poclbm to new branch optimisation as well.
210 2011-08-19 06:19:04 <Delme> Anyone seen conman around?
211 2011-08-19 06:42:52 <Graet> Delme i do often either in -mining or #ozcoin
212 2011-08-19 06:53:07 <shadders> luke-jr:  endian swap algo... take last n bytes put at start of new string in same order.  append previous n bytes from source string to new string etc...
213 2011-08-19 06:53:16 <shadders> that right?
214 2011-08-19 06:53:59 <vegard> that sounds horrible :-S
215 2011-08-19 06:54:55 <shadders> if so n for target is 16 bits to turn diff1 LE into diff1 BE?
216 2011-08-19 06:55:06 <shadders> vegard: it is horrible
217 2011-08-19 06:56:12 <shadders> I thought was supposed to be 32
218 2011-08-19 07:04:02 <lfm> life the universe and everything minus 10?
219 2011-08-19 07:04:23 <shadders> exactly...
220 2011-08-19 07:05:01 <shadders> The only other way I can think to turn diff1 LE into diff1 BE with 32bit chunk size is:
221 2011-08-19 07:05:26 <lfm> whats diff1?
222 2011-08-19 07:06:03 <shadders> take last n byte, reverse and append to new string, repeat for previous n bytes etc... but you're just reversing the string
223 2011-08-19 07:06:33 <shadders> <luke-jr> diff1 LE = 0000000000000000000000000000000000000000000000000000ffff00000000
224 2011-08-19 07:06:34 <shadders> <luke-jr> diff1 xE = 000000000000000000000000000000000000000000000000ffff000000000000\n1576849
225 2011-08-19 07:07:47 <lfm> ya not enuf info, try a different example. what is this from? is it like the target from getwork?
226 2011-08-19 07:08:01 <shadders> yes
227 2011-08-19 07:08:35 <lfm> ok current traget from bitcoind getwork is :
228 2011-08-19 07:08:36 <lfm> "target" : "0000000000000000000000000000000000000000000000864a09000000000000"
229 2011-08-19 07:08:51 <shadders> It's an infuriating example because of the char being grouped in blocks of 4 identicals... can't see if they're reversed or not
230 2011-08-19 07:10:02 <lfm> as a proper hexadecimal number it would be 00000000000094a860000000000000000000000000000000000000000000000
231 2011-08-19 07:10:27 <lfm> does that help?
232 2011-08-19 07:10:51 <shadders> yep coz now I can actually see what's changed...
233 2011-08-19 07:10:56 <shadders> thx...
234 2011-08-19 07:11:28 <lfm> I maybe have the wrong number of zeros there somewhere
235 2011-08-19 07:12:37 <lfm> so ya, convert it to bytes and swap the bytes individulally end for end
236 2011-08-19 07:14:27 <shadders> abcdef -> efcdab?
237 2011-08-19 07:15:38 <lfm> abcdefgh -> ghefcdab ya
238 2011-08-19 07:15:52 <lfm> there should be an even number of bytes
239 2011-08-19 07:16:35 <asher^> i find it easier converting to bin then just reversing and converting back
240 2011-08-19 07:16:44 <lfm> prolly ya
241 2011-08-19 07:17:15 <asher^> otherwise you have to write up a function to split it into 2 character chunks
242 2011-08-19 07:17:37 <lfm> the nasty thing is there may be some different byteswap orders needed in some places, for stuff other than targets
243 2011-08-19 07:17:43 <shadders> it look like each byte has just had it's 4bit halves swapped
244 2011-08-19 07:18:10 <lfm> shadders: nope the bytes do not have their haves swapped. the bytes remain as is
245 2011-08-19 07:18:21 <lfm> just the order of the bytyes changed
246 2011-08-19 07:18:22 <asher^> i was struggling with this the other day for getworks. got some help from the guys here and eventually got it working
247 2011-08-19 07:18:33 <shadders> #bitcoin-endian-victim-support
248 2011-08-19 07:18:40 <lfm> hehe
249 2011-08-19 07:18:50 <asher^> for some reason the previous hash in a block header is LE but the rest is BE
250 2011-08-19 07:18:53 <asher^> confusing as hell
251 2011-08-19 07:18:58 <asher^> for someone like me, anyway
252 2011-08-19 07:19:17 <shadders> oh yeah... brain-hurt
253 2011-08-19 07:19:47 <asher^> i was just writing a function to do a hash
254 2011-08-19 07:19:49 <shadders> When I was in my 20s this wouldn't have been a problem.
255 2011-08-19 07:19:54 <asher^> took me like 90 minutes
256 2011-08-19 07:19:58 <shadders> These days brain is a bit slower
257 2011-08-19 07:20:17 <asher^> for what is essentially bout 6 lines of code
258 2011-08-19 07:20:39 <lfm> ya getwork does some nutty byteswapping cuz they think its helpping but its really not. It allowed some of the old sha256 to work directly with the data. most of bitcoin is littleendian but sha256 is bigendian standard.
259 2011-08-19 07:20:51 <shadders> tell me about it... I spent the best part of day trying to turn solution string into a parseable byte array for bitcoinj
260 2011-08-19 07:21:19 <asher^> im glad i have it done now though
261 2011-08-19 07:21:35 <asher^> there are versions of the algorithm in a couple of languages on the board, but i was doing it in php
262 2011-08-19 07:21:47 <shadders> In the end my solution was... Write a script to iterate through every possible variation of byte swaps, reverses, offsets etc... until one came out that matched the input string...
263 2011-08-19 07:22:13 <asher^> its simple once you know how, as long as you dont think about it
264 2011-08-19 07:22:23 <asher^> when you start thinking about it you get confused as hell
265 2011-08-19 07:22:28 <lfm> hehe shadders, sorry bout that. there are a few people who have been thru this that should have been able to help you
266 2011-08-19 07:22:56 <shadders> was before I found irc...
267 2011-08-19 07:23:33 <shadders> is ok.. my algo found me the right algo in the end... I just had nfi why it worked, just that it did...
268 2011-08-19 07:24:50 <asher^> yeah same here
269 2011-08-19 07:25:00 <asher^> i still dont understand why the stuff is swapped how it is
270 2011-08-19 07:25:56 <lfm> asher^ do you know how sha256 internals use 32 bit adds? It uses the data in big endian order. Thats kinda the orginal reason
271 2011-08-19 07:26:25 <asher^> yeah i figured it was to do with that
272 2011-08-19 07:26:48 <asher^> but the chopping up of the 32b parts of the block header and reversing the parts confused me
273 2011-08-19 07:27:51 <lfm> asher^ then you do about 6 months of open source people optimizing the miner in bitcoind and you get a stripped down sha256 that tries to avoid some of the byte swapping since it does 2 sha256 ops, you get the picture?
274 2011-08-19 07:28:54 <lfm> ie you don't really have much hope of ever following all the rationalization ...
275 2011-08-19 07:29:05 <asher^> yeah i just accept it :)
276 2011-08-19 07:29:33 <shadders> lfm how well document is the bitcoin code?  never had more than passing glance since I'm not a C programmer
277 2011-08-19 07:30:15 <lfm> well even C programmers have been known to throw up their hands at the code since it does a LOT of C++ stuuf! grin
278 2011-08-19 07:31:56 <lfm> seriously tho, it could still be a lot worse. there is someplaces where you follow it thru and are amazed the Satoshi figured it out 2.5 years ago or more and its still working rather well really
279 2011-08-19 07:39:03 <lfm> shadders: so Id say the code is not really very well documented but the code itself is generally quite good and clear that you can read it if you are pretty well up to speed with C and C++ (not just C)
280 2011-08-19 07:44:07 <UukGoblin> more importantly, it works
281 2011-08-19 07:44:25 <UukGoblin> after a year without much of satoshi's intervention
282 2011-08-19 07:45:29 <lfm> I think that that has not been generally accepted as a sole arbiter of good code since the "structured programming" paradyme was adopted back in the '70s
283 2011-08-19 07:46:31 <UukGoblin> oh, you're talking about the code itself... sorry I kinda jumped on the discussion
284 2011-08-19 07:47:01 <UukGoblin> shadders, read up about self-documenting code ;-]
285 2011-08-19 07:47:33 <lfm> true the fact that we are still working with his base code and havn't just thrown it out for a new start is a testament to the quality of the orginal
286 2011-08-19 07:47:54 <UukGoblin> quality or complexity, actually
287 2011-08-19 07:48:00 <cacheson> lfm: not necessarily
288 2011-08-19 07:48:02 <lfm> both maybe
289 2011-08-19 07:48:15 <shadders> UukGoblin that's why java identifiers are usually half a page long.
290 2011-08-19 07:48:23 <UukGoblin> genjix for instance decided to start from scratch
291 2011-08-19 07:48:36 <cacheson> lfm: http://www.joelonsoftware.com/articles/fog0000000069.html
292 2011-08-19 07:48:49 <cacheson> tl;dr: never start over
293 2011-08-19 07:49:13 <lfm> ya there are projects to start over fresh but they are not central yet Id say
294 2011-08-19 07:49:26 <someone42> part of genjix' motivation may have been to avoid centralisation
295 2011-08-19 07:49:27 <UukGoblin> shadders, yes; however, I don't believe longer is better ;-]
296 2011-08-19 07:50:00 <lfm> I gotta go, till later, bye
297 2011-08-19 07:50:02 <shadders> I know, some (most) go way overboard...
298 2011-08-19 07:50:07 <shadders> cya
299 2011-08-19 07:51:01 <shadders> it is easy to read as long as you've got 2 24" widescreens and split yr IDE across both
300 2011-08-19 08:01:42 <xelister> how many hashes it takes avg. to find block at difficulty 1?
301 2011-08-19 08:03:33 <xelister> 2^48/65535 = 4295032833?
302 2011-08-19 08:03:49 <UukGoblin> xelister, yup, that's correct
303 2011-08-19 08:04:21 <UukGoblin> xelister, I've just looked up my scratchpad notes, and the way to get there was different, but the result is the same
304 2011-08-19 08:04:33 <xelister> k
305 2011-08-19 08:16:13 <shadders> still muddled...
306 2011-08-19 08:16:24 <shadders> diff 1 is: 0000000000000000000000000000000000000000000000000000ffff00000000
307 2011-08-19 08:16:27 <shadders> little endian...
308 2011-08-19 08:16:44 <xelister> shadders: https://en.bitcoin.it/wiki/Difficulty
309 2011-08-19 08:16:45 <shadders> pool target is: ffffffffffffffffffffffffffffffffffffffffffffffffffffffff00000000
310 2011-08-19 08:17:05 <shadders> in theory because diff 1 is too hard for slow miners...
311 2011-08-19 08:17:34 <shadders> but as I calculate it the pool target is diff: 0.9999847412109375
312 2011-08-19 08:17:54 <asher^> that sounds about right
313 2011-08-19 08:18:23 <asher^> theres little difference between the two
314 2011-08-19 08:18:26 <shadders> so how can we then say that slow miners are going to get more shares...
315 2011-08-19 08:18:48 <shadders> slightly more... a fraction of a percent...
316 2011-08-19 08:18:58 <asher^> its a tiny difference
317 2011-08-19 08:19:01 <shadders> but not so anyone would notice...
318 2011-08-19 08:19:05 <asher^> why do you want to change it?>
319 2011-08-19 08:20:50 <shadders> I don't but I was told before that it actually makes a difference... I think I was mistold...
320 2011-08-19 08:21:27 <shadders> unless I've messed up the calc somehow...
321 2011-08-19 08:21:41 <asher^> you could make it easier if you wanted, but it would just up the server load
322 2011-08-19 08:22:21 <shadders> don't want to change it... just trying understand why the pools changed it from diff 1 in the first place
323 2011-08-19 08:26:47 <UukGoblin> at least 6MW is being spent for bitcoin mining... hrm...
324 2011-08-19 08:29:07 <shadders> about 20,000 australian homes worth...
325 2011-08-19 08:29:26 <shadders> mind you we do suck a lot of juice...
326 2011-08-19 08:29:49 <UukGoblin> 0.3kW per home?
327 2011-08-19 08:31:50 <shadders> I use about 50kwhrs/day but my bills are twice the average...
328 2011-08-19 08:32:06 <shadders> so 1kw/home... 6000 homes
329 2011-08-19 08:33:06 <UukGoblin> kWh/day... I don't like that unit ;-P that's ~2kW
330 2011-08-19 08:33:40 <kakobrekla> 6000 homes.. not that much
331 2011-08-19 08:34:30 <shadders> apologies if my unit offended you :p  was calculating backwards from my power bill $ figure
332 2011-08-19 08:36:04 <UukGoblin> lol no offense taken ;-]
333 2011-08-19 08:37:21 <hugolp> What does it mean a negative amount of bitcoins in one account of my wallet?
334 2011-08-19 08:38:01 <UukGoblin> hugolp, a bug?
335 2011-08-19 08:38:12 <shadders> delete the wallet.. bankruptcy without the bad credit rating :D
336 2011-08-19 08:38:45 <cacheson> hugolp: you moved more coins out of the account than it actually had
337 2011-08-19 08:38:55 <cacheson> the total wallet balance should still be non-negative
338 2011-08-19 08:39:13 <hugolp> cacheson: the balance of the wallet is 0 indeed
339 2011-08-19 08:39:36 <hugolp> but I did not move intentionally coins from one account to another, I just sent the coins with sendcoins
340 2011-08-19 08:39:38 <hugolp> http://pastebin.com/GYUS2EsL
341 2011-08-19 08:39:56 <UukGoblin> oh, that's stuff is buggy indeed
342 2011-08-19 08:40:01 <UukGoblin> I had that happen to me
343 2011-08-19 08:40:36 <cacheson> hugolp: not sure, never done that myself
344 2011-08-19 08:40:45 <UukGoblin> it's misreported, you don't really have a negative balance
345 2011-08-19 08:41:25 <hugolp> I should mention that I was sending coins from to an address in the other account in the same wallet
346 2011-08-19 08:41:32 <hugolp> *to
347 2011-08-19 08:42:39 <UukGoblin> I was using sendmany in ~0.3.21
348 2011-08-19 08:43:01 <UukGoblin> don't remember exactly what I did, but I had similar issues
349 2011-08-19 08:44:54 <doublec> if you use sendtoaddress then the "" account will be reduced by the amount
350 2011-08-19 08:44:59 <doublec> and can end up with a negative balance
351 2011-08-19 08:49:33 <diki> i've been having this question for a while. But is it possible due to precision errors or stuff like that, a hash <target to be discarded by Phoenix cause it decided it didnt match the difficulty?
352 2011-08-19 08:51:33 <diki> or maybe the other way around, a hash which was pretty close to meeting the current diff, but wasnt <target due to rounding, phoenix to decide it does match the diff
353 2011-08-19 08:51:54 <diki> and send it ?coind thus being rejected by ?coind
354 2011-08-19 08:52:30 <cacheson> diki: there's no rounding, it's all integer operations
355 2011-08-19 08:53:16 <diki> but if you look at the diff it's 1805700.8361937 i.e with a bit more precision
356 2011-08-19 08:53:24 <diki> or does this precision not matter
357 2011-08-19 08:53:36 <cacheson> pretty sure that's fixed-point, not floating point
358 2011-08-19 08:54:16 <diki> uh, define fixed-point
359 2011-08-19 08:54:26 <cacheson> internally represented as an integer
360 2011-08-19 08:54:52 <cacheson> you just put a decimal place at a certain position when showing the number to humans
361 2011-08-19 08:55:44 <zeropointo> "Thank you for providing photo identification. Dwolla support will review the identification for account verification. You will receive an email after a customer service representative has reviewed the account.This process is typically complete within 24 hours."
362 2011-08-19 08:56:05 <zeropointo> hope the person reviewing it doesn't know who peewee herman is lol
363 2011-08-19 08:58:04 <diki> providing fake info?
364 2011-08-19 08:58:09 <diki> goodbye to your money then
365 2011-08-19 08:58:36 <zeropointo> havn't given them any
366 2011-08-19 08:59:25 <zeropointo> like they're going to come to my house at some point and look me in the face? the photo is stupid.
367 2011-08-19 08:59:58 <JFK911> why did dwolla come after you for ID zeropointo ?
368 2011-08-19 09:00:11 <zeropointo> just signing up for an account.
369 2011-08-19 09:00:18 <zeropointo> no idea
370 2011-08-19 09:00:23 <JFK911> oh
371 2011-08-19 09:13:01 <diki> phoenix's askrate is broken
372 2011-08-19 09:13:11 <diki> no matter what you set it at, it always asks for work every 11 seconds
373 2011-08-19 09:13:18 <diki> rarely 12 and 13
374 2011-08-19 09:13:23 <diki> but never what you set it at
375 2011-08-19 09:14:36 <diki> just to point out, i even modified the code of phoenix
376 2011-08-19 09:14:40 <diki> it still asks every 11 seconds
377 2011-08-19 09:14:45 <diki> not 10, 11
378 2011-08-19 09:16:27 <diki> "The power of python"....bogus
379 2011-08-19 09:16:54 <hugolp> back, so then definitively is not a problem that one account has a negative balance, right?
380 2011-08-19 09:17:16 <doublec> right
381 2011-08-19 09:17:50 <hugolp> ok, thanks
382 2011-08-19 09:18:50 <hugolp> another thing, is it posible that the oficial client (with GUI) leaks? After some days the computer it runs on becomes very slow because it has to use the swap disk a lot since there is no memory left
383 2011-08-19 09:19:05 <hugolp> when I close the bitcoin client it gets solved
384 2011-08-19 09:19:10 <hugolp> "solved"
385 2011-08-19 09:19:47 <doublec> it uses a lot of memory
386 2011-08-19 09:19:54 <doublec> depending on the number of connections it has
387 2011-08-19 09:19:59 <doublec> number of transactions kept in memory, etc
388 2011-08-19 09:20:12 <doublec> mine sits aroung 1GB
389 2011-08-19 09:20:15 <hugolp> ahh
390 2011-08-19 09:20:29 <hugolp> that computer has 1GB RAM so that explains a lot
391 2011-08-19 09:20:54 <hugolp> does bitcoind needs so much RAM as well?
392 2011-08-19 09:20:58 <doublec> yes
393 2011-08-19 09:22:51 <diki> i'd be happy if someone told me why
394 2011-08-19 09:23:07 <diki> Crysis 1 (2007) uses less ram
395 2011-08-19 09:23:22 <diki> and it's a game i.e much complex
396 2011-08-19 09:29:16 <hugolp> I want to know why Crysis has become a standard for hardware peformance
397 2011-08-19 09:29:31 <hugolp> But indeed 1GB of RAM for the bitcoin client seems a bit much
398 2011-08-19 09:29:57 <diki> blame it on the coder(s)
399 2011-08-19 09:30:05 <diki> rather, blame it on satoshi
400 2011-08-19 09:30:08 <diki> or there is a memory leak
401 2011-08-19 09:30:21 <kreal-> there is a memory leak
402 2011-08-19 09:30:37 <kreal-> has to allocate 8GB for my bitcoind
403 2011-08-19 09:30:40 <kreal-> had*
404 2011-08-19 09:31:26 <diki> then i dont think wallet encryption or updating to a newer upnpc is the main concern
405 2011-08-19 09:35:59 <hugolp> btw, I just paid for my cat's food in bitcoins. Feels good.
406 2011-08-19 09:38:29 <edcba> who sold you cat's food in bitcoins ??
407 2011-08-19 09:42:09 <hugolp> edcba: there is a internet shop in my country that sells cat and dog food for bitcoins.
408 2011-08-19 09:42:36 <edcba> internet shop selling cat food
409 2011-08-19 09:42:41 <hugolp> My country being in south of Europe under France and next to Portugal (I dont want to give too many clues to not reveal my location...)
410 2011-08-19 09:42:58 <edcba> andorre !
411 2011-08-19 09:43:11 <hugolp> http://www.telepienso.com/
412 2011-08-19 09:43:19 <vegard> < hugolp> My country being in south of Europe under France and next to Portugal (I dont want to give too many clues to not reveal my location...) <-- LOL
413 2011-08-19 09:43:28 <hugolp> andorre is next to Portugal? They have expanded a lot lately then...
414 2011-08-19 09:44:02 <hugolp> ^thats the shop
415 2011-08-19 09:44:24 <kreal-> hugolp I know what you did last summer!
416 2011-08-19 09:48:04 <hugolp> kreal-: for real?
417 2011-08-19 09:52:08 <kreal-> no.
418 2011-08-19 09:52:14 <kreal-> but I know where you live!
419 2011-08-19 12:02:23 <asif> hi
420 2011-08-19 12:43:29 <zamgo> how interesting, there is a graphic in the bicoin-testnet chain
421 2011-08-19 12:43:54 <ThomasV> a graphic ? in ascii ?
422 2011-08-19 12:44:51 <zamgo> strings -n 20 testnet/blk0001.dat
423 2011-08-19 12:45:04 <zamgo> XMP tiff
424 2011-08-19 12:45:07 <zamgo> maytbe
425 2011-08-19 12:47:12 <asher^> any of you guys have .24 working with joelkatz patches?
426 2011-08-19 12:47:46 <jrmithdobbs> what patches are those
427 2011-08-19 12:50:25 <asher^> patches for pool stuff
428 2011-08-19 12:55:07 <CIA-101> bitcoin: Luke Dashjr * r09510add01b6 gentoo/net-p2p/ (4 files in 3 dirs): net-p2p/bitcoind and net-p2p/wxbitcoin: fix boost autodetection for older versions
429 2011-08-19 12:55:18 <Nicksasa> i'm trying to run a 2nd instance of bitcoind so i run it under a diffrent user and change rpcport= & port=
430 2011-08-19 12:55:24 <Nicksasa> but connections stays at 0
431 2011-08-19 12:56:01 <Nicksasa> nevermind, strangely it's downloading the blockchain but "connections" isn't updating, ohwell
432 2011-08-19 13:27:06 <erus`> luke-jr we are moving to haskell
433 2011-08-19 13:27:16 <erus`> so dont worry too much about implementation details
434 2011-08-19 13:28:32 <DukeOfURL> does anyone have a java equivalent of the mtgox_query function in php?
435 2011-08-19 13:28:56 <DukeOfURL> does anyone have a java equivalent of the mtgox_query php function?
436 2011-08-19 13:30:57 <erus`> DukeOfURL: using which library?
437 2011-08-19 13:31:08 <DukeOfURL> android java
438 2011-08-19 13:31:26 <DukeOfURL> i would like to use the mtgox web service from the phone
439 2011-08-19 13:36:07 <asif> hey
440 2011-08-19 13:36:49 <Eliel> DukeOfURL: just use the mobile application mtgox is offering
441 2011-08-19 13:37:38 <DukeOfURL> Eliel: i have used that.  i need an mtgox interface in the app I'm writing.
442 2011-08-19 13:37:55 <Eliel> oh
443 2011-08-19 13:38:28 <DukeOfURL> https://en.bitcoin.it/wiki/MtGox/API
444 2011-08-19 13:40:13 <DukeOfURL> i'd like to use mtgox as the ewallet.  is there a better alternative?
445 2011-08-19 13:44:04 <luke-jr> DukeOfURL: no, that's the best option right now
446 2011-08-19 13:44:13 <luke-jr> for security, at least
447 2011-08-19 13:44:18 <luke-jr> for functionality, WalletBit is nice
448 2011-08-19 13:50:06 <CIA-101> bitcoin: Luke Dashjr * re17497c60689 gentoo/net-p2p/wxbitcoin/ (7 files): net-p2p/wxbitcoin: Need to use || with specific boost slot dependencies, just in case the user has 1.40 and 1.46 :/
449 2011-08-19 13:52:17 <luke-jr> fwiw, the boost 1.46+ issue affects bitcoin-qt as well
450 2011-08-19 13:52:21 <AlonzoTG> Are there any useful IDEs on Linux these days?
451 2011-08-19 13:53:23 <JFK911> yeah use vbox and run visual studio inside that
452 2011-08-19 13:53:48 <AlonzoTG> That's what I thought. =(
453 2011-08-19 13:53:58 <JFK911> i dont like eclipse either
454 2011-08-19 13:54:02 <JFK911> some people do
455 2011-08-19 13:54:10 <AlonzoTG> I don't get eclipse at all.
456 2011-08-19 13:54:11 <JFK911> you can use emacs if you are good
457 2011-08-19 13:54:19 <JFK911> maybe you need more memory heheh
458 2011-08-19 13:54:23 <AlonzoTG> it has a bunch of garbage that gets in your way and is poor interface design from every perspective.
459 2011-08-19 13:54:44 <luke-jr> AlonzoTG: Qt Creator isn't too bad I think?
460 2011-08-19 13:54:50 <JFK911> if you use svn, emacs picks up on it
461 2011-08-19 13:54:54 <luke-jr> I personally use Kate :p
462 2011-08-19 13:54:55 <JFK911> helps a little
463 2011-08-19 13:54:56 <AlonzoTG> And Kdevelop 4 tries to emulate that interface.
464 2011-08-19 13:55:00 <luke-jr> JFK911: who uses svn?
465 2011-08-19 13:55:02 <AlonzoTG> Kdevelop 3 was quite good.
466 2011-08-19 13:55:22 <JFK911> luke-jr: people who want to call their coworker on the telephone and say something liek "Check out revision twelve fifty two"
467 2011-08-19 13:55:36 <luke-jr> JFK911: haha
468 2011-08-19 13:55:43 <luke-jr> JFK911: Bazaar works fine tho ;)
469 2011-08-19 13:55:48 <cacheson> telephone?  what's that?
470 2011-08-19 13:55:57 <luke-jr> cacheson: yeah, we use IM :P
471 2011-08-19 13:56:36 <luke-jr> JFK911: in fact, I'll use Bazaar even if my client has a Subversion repository
472 2011-08-19 13:56:55 <luke-jr> it pretty much just works
473 2011-08-19 13:56:56 <JFK911> how do you talk about git though? "Look what I did just after lunchtime"
474 2011-08-19 13:57:14 <luke-jr> JFK911: you copy and paste commit hashes? :/
475 2011-08-19 13:57:26 <luke-jr> notably, Bazaar has (limited) support for git repos too
476 2011-08-19 13:57:34 <luke-jr> so you *could* use the revno assigned by bzr
477 2011-08-19 13:58:09 <JFK911> im kinda set in my ways tho, im pretty old
478 2011-08-19 13:58:18 <luke-jr> admittedly, Svn does have some killer features
479 2011-08-19 13:58:32 <luke-jr> like copies, metadata, and cherry picking
480 2011-08-19 13:58:33 <JFK911> its alright.  seems light and stable
481 2011-08-19 13:59:08 <JFK911> i see that being connected all the time is a problem for some people
482 2011-08-19 13:59:28 <luke-jr> Bazaar mostly works fine for me if upstream is Svn :P
483 2011-08-19 13:59:30 <luke-jr> including for offline dev
484 2011-08-19 13:59:38 <cjdelisle> git > svn > cvs > hammer and chisel
485 2011-08-19 13:59:47 <JFK911> sccs
486 2011-08-19 13:59:50 <JFK911> heheh
487 2011-08-19 13:59:59 <JFK911> never forget what att gave us
488 2011-08-19 14:00:05 <CIA-101> bitcoin: Luke Dashjr * re529359e62de gentoo/net-p2p/bitcoin-qt/ (Manifest bitcoin-qt-9999.ebuild): net-p2p/bitcoin-qt: Need to use || with specific boost slot dependencies, just in case the user has 1.40 and 1.46 :/
489 2011-08-19 14:00:19 <luke-jr> cjdelisle: bzr, git, and svn all have advantages
490 2011-08-19 14:00:33 <luke-jr> bzr is surprisingly overlooked
491 2011-08-19 14:00:34 <JFK911> people should use whats best for them
492 2011-08-19 14:00:44 <luke-jr> JFK911: that's pretty much bzr's attitude
493 2011-08-19 14:00:48 <cjdelisle> advantage of svn is noone will ever fork your project... ever
494 2011-08-19 14:00:49 <JFK911> i just have a stash for personal projects
495 2011-08-19 14:00:53 <luke-jr> cjdelisle: why not?
496 2011-08-19 14:00:54 <JFK911> not a large group of people
497 2011-08-19 14:01:01 <luke-jr> cjdelisle: bzr can fork svn upstream no problem
498 2011-08-19 14:01:10 <JFK911> yeah why not?  worked between my intern and i
499 2011-08-19 14:01:15 <luke-jr> bzr branch svn://& <-- makes a complete clone, including history
500 2011-08-19 14:01:16 <JFK911> now he didnt LIKE it
501 2011-08-19 14:01:19 <cjdelisle> I don't know bzr but I understand it is better than cvsvn
502 2011-08-19 14:01:27 <JFK911> but that helped him learn about teamwork
503 2011-08-19 14:01:48 <luke-jr> cjdelisle: bzr can treat svn as just another DVCS branch
504 2011-08-19 14:52:08 <diki> was going to ask
505 2011-08-19 14:52:26 <diki> but can two bitcoin addresses which are identical exist, but their private keys to be different?
506 2011-08-19 14:53:51 <diki> basically, since it is possible to make custom addresses, what is stopping people from making identical addresses as another person's wallet addr?
507 2011-08-19 14:54:07 <edcba> lolll
508 2011-08-19 14:54:50 <diki> what is so funny?
509 2011-08-19 14:55:37 <edcba> not that easy
510 2011-08-19 14:55:45 <diki> not that easy what?
511 2011-08-19 14:55:49 <log0s> diki: it's possible (due to hash collisions) but extremely unlikely
512 2011-08-19 14:56:11 <diki> so can two addresses which are identical exist but have different private keys?
513 2011-08-19 14:56:31 <ThomasV> diki: in theory yes. in practice, NO
514 2011-08-19 14:56:34 <log0s> diki: they *can*, but in practice it will probably never happen
515 2011-08-19 14:56:59 <diki> so you are saying if someone right now either by luck or by large processing power generated the same address as X person
516 2011-08-19 14:57:15 <diki> they can basically receive anything that addr receives?
517 2011-08-19 14:57:45 <ThomasV> by luck ?
518 2011-08-19 14:57:50 <Eliel> yes, if they happen to generate the same address, they can use the coins how they please.
519 2011-08-19 14:58:00 <diki> "Vanitygen uses the OpenSSL random number generator.  "
520 2011-08-19 14:58:05 <diki> i.e its all random
521 2011-08-19 14:58:29 <diki> so i thought that there is a fair amount of luck in this
522 2011-08-19 14:59:09 <luke-jr> diki: you're more likely to have a block hash collision
523 2011-08-19 14:59:19 <ThomasV> diki: the chances that you die from a meteor are way higher
524 2011-08-19 14:59:23 <luke-jr> which would basically destroy Bitcoin
525 2011-08-19 14:59:38 <Eliel> luke-jr: you sure?
526 2011-08-19 14:59:49 <luke-jr> Eliel: ?
527 2011-08-19 15:00:46 <Eliel> I mean, are you sure it would be a problem if two block hashes happened to coincide?
528 2011-08-19 15:00:59 <log0s> a single block hash collision will not destroy bitcoin...someone finding a flaw in sha256 making it easy to find collisions could destroy bitcoin
529 2011-08-19 15:01:03 <phantomcircuit> Eliel, yes it would b3e
530 2011-08-19 15:01:23 <phantomcircuit> a single block hash collision would cause the block chain to be a circle
531 2011-08-19 15:02:02 <Eliel> phantomcircuit: that sounds like an easy to prevent situation. Simply reject new blocks with the same hash as an older one.
532 2011-08-19 15:02:30 <phantomcircuit> Eliel, it's so unlikely to happen
533 2011-08-19 15:02:37 <phantomcircuit> like
534 2011-08-19 15:02:45 <phantomcircuit> heat death of the universe unlikely
535 2011-08-19 15:02:48 <kreal-> Rate My Bitcoin Rig: http://opencode.dk/
536 2011-08-19 15:04:10 <Eliel> phantomcircuit: so, what's the average time between the network finding two blocks with the same hash?
537 2011-08-19 15:04:30 <phantomcircuit> Eliel, a number approximating infinity
538 2011-08-19 15:04:44 <Eliel> phantomcircuit: there is no number approximating infinity
539 2011-08-19 15:04:58 <phantomcircuit> a number which approximates infinity on a human time scale
540 2011-08-19 15:05:04 <CIA-101> bitcoin: Luke Dashjr * r5133ba9dfbbe gentoo/net-p2p/wxbitcoin/ (8 files in 2 dirs): net-p2p/wxbitcoin: Apply boost fs v3 patch to support newer boost versions
541 2011-08-19 15:05:16 <phantomcircuit> Eliel, blocks are hashed with sha256
542 2011-08-19 15:05:23 <phantomcircuit> so there are 2^256 possible block hashes
543 2011-08-19 15:05:27 <ThomasV> Eliel: is there enough matter in the universe to store the blockchain until that happens ?
544 2011-08-19 15:05:31 <Dagger3> presumably it's something like 2^256 * 10 minutes
545 2011-08-19 15:05:53 <phantomcircuit> so yeah Dagger3 is right
546 2011-08-19 15:06:00 <Eliel> Dagger3: more like 2^128 * 10 minutes ... although the difficulty reduces it too.
547 2011-08-19 15:07:11 <Dagger3> even with 2^128 it's still 10^41 seconds
548 2011-08-19 15:07:12 <diki> so does ocl vanity gen have some kind of protection against finding collisions?
549 2011-08-19 15:07:28 <phantomcircuit> Eliel, it's 2^255 * 10 minutes
550 2011-08-19 15:07:30 <Dagger3> when the age of the universe is 10^17 seconds, it's not worth worrying about
551 2011-08-19 15:07:44 <phantomcircuit> diki, no it's just absurdly unlikely
552 2011-08-19 15:08:04 <phantomcircuit> Eliel, 1.10079248 ?? 10^69 millenia good enough for you?
553 2011-08-19 15:08:06 <diki> by unlikely u mean it would take so much time that there'd be no point trying
554 2011-08-19 15:08:19 <diki> but otherwise it's 100% possible?
555 2011-08-19 15:08:27 <phantomcircuit> facepalm
556 2011-08-19 15:08:33 <Eliel> diki: well, by all intents and purposes, every miner is already trying.
557 2011-08-19 15:08:33 <phantomcircuit> diki, lrn2crypto
558 2011-08-19 15:08:49 <diki> i am not talking about the stupid blocks
559 2011-08-19 15:08:57 <diki> i was talking about the addresses
560 2011-08-19 15:09:02 <diki> wallet addr
561 2011-08-19 15:09:10 <ThomasV> diki: itz da same
562 2011-08-19 15:09:43 <Eliel> diki: google birthday attack.
563 2011-08-19 15:13:23 <log0s> for addresses it would require finding a private key which has a public key that when hashed by sha256, and then hashed by ripemd160, creates a collision with the ripemd160 hash of the sha256 hash of the public key of a different private key...
564 2011-08-19 15:13:26 <diki> how i understand the word unlikely is that it can happen, but it will take so much time it isnt even worth trying
565 2011-08-19 15:13:38 <diki> thus making you use the word unlikely
566 2011-08-19 15:14:01 <Eliel> diki: it's that unlikely, basically.
567 2011-08-19 15:17:30 <diki> log0s:so it's not as simple as just entering the address you want in ocl vanity gen and even if a priv key is generated there are other things that need to match in order for this to work?
568 2011-08-19 15:18:11 <mtrlt> no, all you need is the private key
569 2011-08-19 15:18:23 <diki> then why did log0s even post all that
570 2011-08-19 15:18:42 <mtrlt> the private key that results in the address you want.
571 2011-08-19 15:19:18 <diki> then isnt it time to ban vanity gen?
572 2011-08-19 15:19:24 <mtrlt> how do you ban it
573 2011-08-19 15:19:35 <diki> that i dont know
574 2011-08-19 15:19:38 <cacheson> you need a hash collision in order to get a different private key but the same bitcoin address
575 2011-08-19 15:19:38 <mtrlt> if finding a collision was that easy, bitcoin would be inherently broken
576 2011-08-19 15:19:40 <diki> why did gaving even write that
577 2011-08-19 15:20:03 <cacheson> I'd guess you're more likely to just generate the same private key as someone else
578 2011-08-19 15:20:07 <mtrlt> finding a collision like that with a specific address would take like 2^160 trie
579 2011-08-19 15:20:08 <mtrlt> or so
580 2011-08-19 15:20:12 <cacheson> and that is... "unlikely"
581 2011-08-19 15:20:18 <mtrlt> cacheson: no
582 2011-08-19 15:20:26 <mtrlt> cacheson: there are 2^256 private keys but 2^160 addresses.
583 2011-08-19 15:20:35 <mtrlt> on average, there are 2^96 private keys for each address
584 2011-08-19 15:20:47 <enquirer> couldn't block chain be used for any type of important timestamping, not only money?
585 2011-08-19 15:21:17 <cacheson> mtrlt: ah
586 2011-08-19 15:21:27 <enquirer> contracts for example
587 2011-08-19 15:22:06 <mtrlt> yep there can be lots of creative uses of the block chain :p
588 2011-08-19 15:22:19 <diki> so, the way vanity gen works, is it possible for people to make pools to generate addresses?
589 2011-08-19 15:22:29 <log0s> enquirer: it *could* be, but the bitcoin developers want the blockchain to be as useless as possible
590 2011-08-19 15:22:30 <mtrlt> yes of course
591 2011-08-19 15:22:47 <mtrlt> but it won't generate a collision before the universe is dead
592 2011-08-19 15:22:50 <mtrlt> a single one.
593 2011-08-19 15:23:18 <diki> by 2100 there will be 9-10 billion people on earth
594 2011-08-19 15:23:24 <diki> thus more people with PCs
595 2011-08-19 15:23:28 <diki> thus more GPUs
596 2011-08-19 15:23:29 <mtrlt> and?
597 2011-08-19 15:24:03 <luke-jr> that sounds like a low estimate.
598 2011-08-19 15:24:23 <mtrlt> i doubt we'll have GPUs in 2100 :P
599 2011-08-19 15:24:49 <diki> sure
600 2011-08-19 15:24:55 <diki> its possible for them to be replaced
601 2011-08-19 15:25:04 <mtrlt> but assuming moore's law.
602 2011-08-19 15:25:56 <diki> by 2100 the max 450/70 mh/s a card can do right now will prolly be a laughable amount in the future
603 2011-08-19 15:26:23 <luke-jr> by 2100, Bitcoin as we know it will be dead
604 2011-08-19 15:26:54 <enquirer> maybe it turns out p=np by 2100
605 2011-08-19 15:27:35 <enquirer> and all cryptography goes down the drain
606 2011-08-19 15:30:50 <asif> hi
607 2011-08-19 15:30:54 <luke-jr> o hai
608 2011-08-19 15:31:03 <mtrlt> enquirer: all?
609 2011-08-19 15:31:08 <diki> wanna be friends?
610 2011-08-19 15:31:47 <diki> so i use around 2-3 fingers when typing
611 2011-08-19 15:31:54 <diki> how do i use all 10?
612 2011-08-19 15:32:37 <diki> the qwerty keyboard should allow people to use at least 10, but i cant
613 2011-08-19 15:32:43 <diki> i can only use 4 max
614 2011-08-19 15:34:53 <log0s> interesting: http://packetstormsecurity.org/news/view/19719/AES-Crypto-Compromised-By-Groundbreaking-Attack.html
615 2011-08-19 15:35:08 <log0s> the encrypted wallets should still be fine, though
616 2011-08-19 15:36:05 <diki> heheheheheh
617 2011-08-19 15:36:10 <luke-jr> diki: are you stupid?
618 2011-08-19 15:36:11 <diki> seems people are one step closer
619 2011-08-19 15:36:23 <diki> luke-jr:i expect you to take those words back
620 2011-08-19 15:36:35 <luke-jr> diki: QWERTY was designed for one purpose: to make typing more difficult
621 2011-08-19 15:36:45 <log0s> trillions of years to crack a key is still a long time
622 2011-08-19 15:37:14 <diki> luke-jr:so?
623 2011-08-19 15:37:20 <diki> my question was totally different
624 2011-08-19 15:38:43 <asif> hi
625 2011-08-19 15:38:53 <diki> ciao
626 2011-08-19 15:49:26 <luke-jr> so who's gonna redo wallet encryption with PGP? :P
627 2011-08-19 15:50:17 <D0han> person with key
628 2011-08-19 15:51:00 <samr7> luke-jr, will it be better off with pgp?
629 2011-08-19 15:51:09 <luke-jr> samr7: afaik PGP isn't broken
630 2011-08-19 15:51:31 <samr7> luke-jr, what's wrong with the current impl?
631 2011-08-19 15:51:40 <luke-jr> it uses AES
632 2011-08-19 15:51:47 <luke-jr> which was just broken
633 2011-08-19 15:52:06 <samr7> so somebody figures out a 3x improvement
634 2011-08-19 15:52:12 <samr7> that's not breaking
635 2011-08-19 15:52:15 <luke-jr> yes it is
636 2011-08-19 15:52:30 <samr7> breaking is a sub-exponential improvement
637 2011-08-19 15:52:40 <riush> 2**126 still takes /some/ time...
638 2011-08-19 15:52:49 <luke-jr> broken tends to snowball
639 2011-08-19 15:52:55 <cacheson> 
640 2011-08-19 15:52:58 <phantomcircuit> luke-jr, https://github.com/bitcoin/bitcoin/pull/232#issuecomment-1395236
641 2011-08-19 15:53:03 <luke-jr> using AES now is like playing with fire :P
642 2011-08-19 15:53:17 <samr7> luke-jr, actually AES256 that they use is a bit broken, there was a published attack a while back
643 2011-08-19 15:53:19 <phantomcircuit> for the lulz
644 2011-08-19 15:53:44 <phantomcircuit> AES256 is LESS secure than 128
645 2011-08-19 15:53:53 <samr7> ^^ indeed!
646 2011-08-19 15:53:57 <phantomcircuit> they use a slightly different key schedule
647 2011-08-19 15:54:14 <phantomcircuit> i tried to warn people about that when the wallet crypto was being implemented
648 2011-08-19 15:54:17 <phantomcircuit> but people ignored me
649 2011-08-19 15:54:19 <samr7> the wallet encryption also uses PKCS#5 1.5
650 2011-08-19 15:54:20 <marf_away> doees it help at mining?
651 2011-08-19 15:54:20 <phantomcircuit> meh
652 2011-08-19 15:54:25 <copumpkin> marf_away: no
653 2011-08-19 15:54:57 <luke-jr> phantomcircuit: what's more fun, is that the 1) passphrase isn't mlock'd, 2) mlock fails on any standard Linux install, and 3) bitcoind ignores the mlock failure and goes on as if it was fine
654 2011-08-19 15:55:17 <luke-jr> marf_away: mining is SHA-256
655 2011-08-19 15:55:23 <vegard> phantomcircuit> AES256 is LESS secure than 128 ?
656 2011-08-19 15:55:25 <luke-jr> marf_away: if that is broken, Bitcoin falls apart completely
657 2011-08-19 15:56:08 <samr7> luke-jr, so there's a slim chance of sensitive information being leaked to swap
658 2011-08-19 15:56:16 <phantomcircuit> vegard, correct
659 2011-08-19 15:56:17 <luke-jr> samr7: yep
660 2011-08-19 15:56:31 <luke-jr> samr7: most important IMO is the false sense of security people have thinking it's mlock'd when it isn't
661 2011-08-19 15:56:43 <phantomcircuit> vegard, they are *not* the same algorithm they are merely very similar algorithms
662 2011-08-19 15:57:39 <samr7> luke-jr, people here = who?
663 2011-08-19 15:58:01 <vegard> phantomcircuit: doesn't it have twice as many bits in the key?
664 2011-08-19 15:58:02 <luke-jr> samr7: anyone who glances at the code and sees mlock
665 2011-08-19 15:58:43 <phantomcircuit> vegard, yes but the actual algorithm used by aes 128 is less secure than the one used by aes 256
666 2011-08-19 15:58:49 <phantomcircuit> it is not merely a question of key length
667 2011-08-19 15:59:20 <vegard> this is not about the recently published attack, is it?
668 2011-08-19 15:59:26 <phantomcircuit> no it's not
669 2011-08-19 15:59:31 <samr7> luke-jr, so should they get rid of mlock, or fix the behavior?
670 2011-08-19 15:59:41 <phantomcircuit> it has been known for a long time that aes 128 is more secure than aes 256
671 2011-08-19 15:59:46 <luke-jr> samr7: either use mlock or don't, but don't have it in there doing nothing
672 2011-08-19 15:59:49 <vegard> so how many bits of security for each of them?
673 2011-08-19 16:02:44 <vegard> 
674 2011-08-19 16:02:53 <vegard> wikipedia. I assume that's it?
675 2011-08-19 16:05:20 <samr7> luke-jr, I agree, the implementation of that does look kinda messy
676 2011-08-19 16:05:42 <samr7> it has caveats all over the place about other ways that passwords can be leaked to swap
677 2011-08-19 16:07:05 <iddo> vegard: ECDSA that bitcoin uses for signing txns doesn't have more than 128 bits of security, so you can argue that aes128 is enough
678 2011-08-19 16:07:08 <samr7> I can't see it helping an attacker other than maybe a law enforcement guy who is clever enough to look at the swap on a seized machine
679 2011-08-19 16:07:54 <iddo> anyway the aes256 practical attacks are related-key attacks, not relevant when using single key to encrypt wallet
680 2011-08-19 16:09:28 <vegard> samr7: virtual machines maybe
681 2011-08-19 16:10:10 <samr7> vegard, true!
682 2011-08-19 16:18:55 <Eliel> iddo: huh? I thought the ECDSA used was 256 bit?
683 2011-08-19 16:19:46 <samr7> Eliel, it is, but EC methods are considered to provide N/2 "bits" of equivalent security to a secret key